A Sensorless Control Method for Permanent Magnet Synchronous Motor with Random Pulse Sequence Injection
By adopting a random pulse sequence voltage signal generator and signal demodulation strategy in the permanent magnet synchronous motor, the electromagnetic interference and noise problems generated by the high-frequency signal injection method in the permanent magnet synchronous motor are solved, and high-precision rotor position estimation and control are achieved.
Patent Information
- Application Number
- CN202211628494.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-12-17
- Publication Date
- 2025-07-11
- Estimated Expiration
- 2042-12-17
AI Technical Summary
The existing high-frequency signal injection method generates strong electromagnetic interference and harsh noise in the position sensor control of permanent magnet synchronous motors, limiting its application in actual industrial fields.
The random pulse sequence voltage signal generator is superimposed on the d-axis current loop output of the permanent magnet synchronous motor, and the envelope of the high-frequency response signal is extracted through the Clark transformation and signal demodulation strategy, and the rotor position and rotation speed are obtained in combination with the orthogonal phase-locked loop processing.
The audible noise problem of high-frequency signal injection method is significantly suppressed, the rotor position estimation accuracy is improved, and the application field of positionless sensor control method is broadened.
Smart Images

Figure CN116232135B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of motor control, and particularly to a sensorless control method for a permanent magnet synchronous motor by injecting a random pulse sequence. Background Art
[0002] In recent years, due to the advantages of high power density, high efficiency, excellent speed regulation performance and small volume, permanent magnet synchronous motors have been widely used in various high-performance motor drive systems. High-performance control of permanent magnet synchronous motors usually requires the installation of position sensors to obtain motor speed and rotor position information. However, the installation of mechanical position sensors will increase the system volume and cost and reduce the system robustness. In the past 20 years, various sensorless control methods for permanent magnet synchronous motors have been proposed, which can be mainly divided into two types: the fundamental frequency model method applicable to the medium and high speed range and the high-frequency injection method applicable to the zero and low speed range. The fundamental frequency model method mainly obtains the back electromotive force or stator magnetic flux of the motor by constructing an observer, and then calculates the rotor position of the motor. However, due to the low signal-to-noise ratio of the back electromotive force or stator magnetic flux in the zero and low speed range, it cannot ensure its normal operation in the zero and low speed range.
[0003] The high-frequency injection method based on salient pole tracking can overcome the limitations of the fundamental frequency model method in the zero and low speed range. It instantaneously excites the permanent magnet synchronous motor by injecting a high-frequency voltage or current signal to generate a high-frequency response containing position information. By designing a corresponding signal demodulation strategy to extract and process the high-frequency response, the motor speed and position information can be obtained. According to the different injection signals, the high-frequency injection method mainly includes: the high-frequency sine signal injection method, the high-frequency square wave signal injection method and the high-frequency pulse signal injection method. Compared with the high-frequency sine signal injection, the high-frequency square wave signal and high-frequency pulse signal injections are easier to be digitally implemented and have a higher injection frequency, which is beneficial to the design of the signal demodulation link and the improvement of the system bandwidth. However, due to the high injection signal frequency, the excited high-frequency current response will cause strong electromagnetic interference and harsh noise. The strong electromagnetic interference may affect the normal operation of the equipment and even make it unable to work. The harsh noise is not only unfavorable to the environment but also may affect human health, which will undoubtedly limit the application of such methods in the actual industrial field.
[0004] Therefore, to ensure the practicability of the high-frequency signal injection method, it has important theoretical significance and practical value to realize sensorless control of a permanent magnet synchronous motor with low noise on the premise of ensuring the rotor position estimation accuracy. Summary of the Invention
[0005] To solve the problem that strong electromagnetic interference and harsh noise will be generated when the signal injection method is used to realize sensorless control of a permanent magnet synchronous motor in the zero and low speed range in the prior art, the present invention provides a sensorless control method for a permanent magnet synchronous motor based on injecting a random pulse sequence.
[0006] A sensorless control method for a permanent magnet synchronous motor with random pulse sequence injection, the steps are as follows:
[0007] Step 1: Use a random pulse sequence voltage signal generator to superimpose a voltage signal with a random pulse sequence on the output of the d-axis current loop estimated by the vector control of the permanent magnet synchronous motor.
[0008] Step 2: Obtain the three-phase current of the permanent magnet synchronous motor through A / D sampling, and convert the three-phase current from the natural coordinate system to the stationary coordinate system through Clark transformation.
[0009] Step 3: Separate the fundamental wave signal and the high-frequency response signal through a signal demodulation strategy, and extract the envelope of the high-frequency response signal to obtain an orthogonal signal related to the rotor position.
[0010] Step 4: Use an orthogonal phase-locked loop to process the orthogonal signal related to the rotor position to obtain the estimated rotor speed of the motor and position
[0011] Step 5: Use the estimated motor speed and position for speed closed-loop control and Park and inverse Park coordinate transformations respectively to achieve sensorless closed-loop control of the permanent magnet synchronous motor.
[0012] According to the sensorless control method for a permanent magnet synchronous motor with random pulse sequence injection of the present invention, the sequence of the injected pulse voltage signal is randomly selected and generated by a random pulse sequence voltage signal generator.
[0013] According to the sensorless control method for a permanent magnet synchronous motor with random pulse sequence injection of the present invention, the pulse sequence of the injected pulse voltage in step 1 includes: sequence a: “++--00”, sequence b: “--++00”, sequence c: “+-0-+0” and sequence d: “-+0+-0”. Here, the four different pulse sequences are all composed of six states, and each state lasts for one PWM carrier cycle. “+” represents positive voltage injection, “-” represents negative voltage injection, and “0” represents zero voltage injection.
[0014] According to the sensorless control method for a permanent magnet synchronous motor with random pulse sequence injection of the present invention, the injected pulse voltage signal in the dq coordinate system is expressed as:
[0015]
[0016] where, u dh and u qh are the injected voltage signals on the d-axis and q-axis respectively, and u injThe voltage signals with different pulse sequences after random selection, T p1 is the injection period of the pulsed voltage signal, u s_x (x = a, b, c, d) is the sequence sample of the injected pulse signal, t is the time, and n is the number of periods of the injected pulsed voltage.
[0017] The sequence sample of the injected pulse signal can be expressed as:
[0018] u s_x (t, T p1 ) = R[u s_a (t, T p1 ), u s_b (t, T p1 ), u s_c (t, T p1 ), u s_d (t, T p1 )]
[0019] where R[] is a random operator, and the pulsed voltage signals with sequences a, b, c, d are respectively expressed as:
[0020]
[0021] u s_b (t, T p1 ) = -u s_a (t, T p1 )
[0022]
[0023] u s_d (t, T p1 ) = -u s_c (t, T p1 )
[0024] where t r (t, T p1) is the remainder of t divided by T p1 , and V inj is the amplitude of the injected voltage. Even if the pulse signals with sequences c or d are injected arbitrarily, the estimated d-axis current will not have a DC offset. However, when the injection probabilities of the pulse signals with sequences a and b are not equal or have non-ideal mathematical expectations due to non-ideal factors, the estimated d-axis current may have a DC offset. To avoid this problem, the pulsed voltage signals of sequences a and b are injected in pairs. Define T p2 = 2T p1 , then the sequence sample of the injected pulsed voltage signal can be rewritten as:
[0025] u s_x (t, T pi)=R[u s_c (t,T p1 ),u s_d (t,T p1 ),u s_c (t,T p2 ) / 2,u s_d (t,T p2 ) / 2]
[0026] Here, i is 1 or 2. That is, the frequency range of the injected pulse voltage signal is f p1 (=1 / T p1) to f p2 (=1 / T p2 ).
[0027] According to the sensorless control method for a permanent magnet synchronous motor injected with a random pulse sequence of the present invention, the signal demodulation method in step 3 is specifically as follows:
[0028] First, the three-phase current of the permanent magnet synchronous motor is obtained by A / D sampling, and the three-phase current is converted from the natural coordinate system to the stationary coordinate system by Clark transformation. Since the injection signal frequency is much higher than the motor operating frequency, it can be considered that the fundamental current remains unchanged within an injection cycle. αβ The two current differences at the demodulation moments of adjacent signals are subtracted to achieve signal demodulation. First, the corresponding demodulation signal sym is designed according to the different sequences of the pulse signal:
[0029]
[0030] sym s_x (t,T pi )=R[sym s_c (t,T p1 ),sym s_d (t,T p1 ),sym s_c (t,T p2 ),sym s_d (t,T p2 )]
[0031] Here, the demodulated signal sym s_c (t,T pi) With sym s_d (t,T pi) It can be expressed as:
[0032]
[0033] sym s_d (t,T pi )=-sym s_c (t,T pi )
[0034] The estimated high-frequency response current on the d-axis excited by the injection pulse signal can be expressed as:
[0035]
[0036] i s_x (t,T pi ) = R[i s_a (t,T p1 ),i s_b (t,T p1 ),i s_a (t,T p2 ),i s_b (t,T p2 )]
[0037] Here, i dh is the d-axis current response signal, i inj is the current response signal excited by the voltage signal with different pulse sequences after random selection, and i s_x (x = c, d) is the sequence sample of the injected current response signal. The current response sample can be expressed as:
[0038]
[0039] i s_d (t,T pi ) = -i s_c (t,T pi )
[0040] Here, k I is the amplitude of the current response signal, and the envelope of the high-frequency current response in the stationary coordinate system can be extracted:
[0041]
[0042] Here, Δi cos and Δi sin are orthogonal signals containing position information, and i α,β0 , i α,β1 and i α,β2 are the current sampling values at the signal demodulation times t r (t,T pi) is 0, T pi / 6, and T pi / 3, or the signal demodulation time t r (t,T pi) is T pi / 2, 2T pi / 3, and 5T pi / 6, ΔT is the signal demodulation period, and η is a variable that can be expressed as
[0043] Since signal demodulation is performed 1 or 2 times in an injection period, a low-pass filter with a high cut-off frequency is used to achieve the difference of the quadrature signals within a control period, where the low-pass filter is high enough not to cause a phase change of the quadrature signals, and the filtered quadrature signal Δi cos1 and Δi sin1 can be expressed as:
[0044]
[0045] After normalization, the unit quadrature signal containing rotor position information can be expressed as:
[0046]
[0047] According to the sensorless control method of a permanent magnet synchronous motor with random pulse sequence injection of the present invention, the equivalent position error ε is obtained by the contrast method:
[0048]
[0049] Furthermore, the input position error is adjusted to 0 through a PI regulator, so that the estimated motor speed and position converge to the actual values, and the estimated motor speed and position can be expressed as
[0050]
[0051] Here, k p and k i are the proportional and integral gains of the phase-locked loop respectively.
[0052] Advantages of the present invention: The present invention uses a random pulse sequence generator to generate a pulse voltage signal with a random pulse sequence and superimposes it on the output of the d-axis current loop estimated by the vector control of the permanent magnet synchronous motor. The pulse samples of the random pulse sequence are generated by a random combination of positive voltage, negative voltage and zero voltage injection; after the pulse voltage signal with a random pulse sequence is injected into the motor stator, by using i αβThe subtraction operation is performed on the current difference between two adjacent signal demodulation instants to extract the high-frequency response signal. It is multiplied by the demodulation signal and then normalized to obtain the unit orthogonal signal containing the rotor position information, and then the rotor position information is extracted through the phase-locked loop processing. In the present invention, the sequence of the injected pulse voltage has extremely strong randomness, so as to ensure that all discrete spectra in the high-frequency current response of the excitation are eliminated, convert the noise spectrum of the excitation current response signal into a continuous spectrum approximated to white noise, and at the same time, the continuous spectrum distribution is also relatively wide, thus significantly suppressing the audible noise problem existing in the sensorless control method based on the high-frequency signal injection method. The experimental results show that under the condition of the permanent magnet synchronous motor with rated load, the peak value of the current response spectrum in the proposed injection method can be reduced by about 25 dB compared with the traditional method. At the same time, the subtraction operation is performed on the current difference between two adjacent signal demodulation instants to extract the high-frequency response signal, which can achieve high-precision rotor position estimation. The experimental results show that when the permanent magnet synchronous motor operates at 300 rpm and under the half-rated load condition, the signal demodulation method adopted has higher precision compared with the traditional method, and the estimation precision is improved by about 3.5°.
[0053] This method has significant audible noise suppression ability and good dynamic and static performance, can broaden the application field of the sensorless control method based on the high-frequency signal injection method, and has important theoretical significance and practical value. Brief Description of the Drawings
[0054] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained according to these drawings without creative efforts.
[0055] Figure 1 It is the control block diagram of the sensorless control method for the permanent magnet synchronous motor with random pulse sequence injection.
[0056] Figure 2 It is the random sequence sample of the injected pulse voltage and the high-frequency current response of the excitation. Figure 2 In (a), (b), (c) and (d) in it, the injection samples of sequences a, b, c and d and the corresponding high-frequency current responses are respectively shown.
[0057] Figure 3 It is the schematic diagram of the high-frequency pulse sequence, demodulation symbol variable, high-frequency current response of the method of the present invention and the current sampling sequence.
[0058] Figure 4It is a comparative analysis diagram of the FFT and PSD of the phase-a current of the method of the present invention and the traditional method when the permanent magnet synchronous motor operates at 300 rpm and rated load conditions; Figure 4 In (a), (b), (c), and (d), they are respectively the experimental results of the 2.5 kHz high-frequency square wave voltage injection method, the 1.25 kHz high-frequency square wave voltage injection method, the 1.67 kHz high-frequency pulse voltage injection method, and the method of the present invention. From top to bottom in each figure are the phase-a current, the current PSD, and the current FFT results.
[0059] Figure 5 It is the signal waveform diagram of the signal demodulation method of the method of the present invention. Figure 5 In (a), from top to bottom are respectively Δi αh , Δi βh , Δi cos_pu and Δi sin_pu , Figure 5 In (b), from top to bottom are respectively the actual position, the estimated position, the position error, and the phase-a current.
[0060] Figure 6 It is the experimental comparison result of the traditional square wave injection method and the sensorless method provided by the present invention. Figure 6 In (a) and (b), from top to bottom are respectively Δi cos_pu , Δi sin_pu , the estimated position, and the position error.
[0061] Figure 7 It is the experimental result of the method of the present invention under the acceleration and deceleration conditions at rated load. Figure 7 In (a), from top to bottom are respectively the actual speed, the estimated speed, the phase-a current, and the speed error, Figure 7 In (b), from top to bottom are respectively the actual position, the estimated position, the phase-a current, and the position error.
[0062] Figure 8 It is the experimental result of the method of the present invention under the forward and reverse running conditions at half rated load. Figure 8 From top to bottom in it are respectively the actual speed, the phase-a current, the estimated position, and the position error.
[0063] Figure 9 It is the experimental result of the method of the present invention under the rated load step condition when the speed is 300 rpm. Figure 9 From top to bottom in it are respectively the actual speed, the estimated speed, the phase-a current, and the position error. Specific embodiments
[0064] To make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and do not limit the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative efforts fall within the scope of protection of the present invention.
[0065] According to Figure 1 the control block diagram of the sensorless control method for a permanent magnet synchronous motor with random pulse sequence injection shown in the figure, the sensorless control method provided by the present invention mainly includes a random pulse sequence voltage signal generator, vector closed-loop control, and a position estimation process. The present invention provides a sensorless control method for a permanent magnet synchronous motor with random pulse sequence injection, and the steps are as follows:
[0066] Step 1: Use a random pulse sequence voltage signal generator to superimpose a voltage signal with a random pulse sequence on the output of the d-axis current loop estimated by the vector control of the permanent magnet synchronous motor. The pulse sequences of the injected pulse voltage include: sequence a: “++--00”, sequence b: “--++00”, sequence c: “+-0-+0” and sequence d: “-+0+-0”. Here, the four different pulse sequences are all composed of six states, and each state lasts for one PWM carrier period. “+” represents positive voltage injection, “-” represents negative voltage injection, and “0” represents zero voltage injection.
[0067] The injected pulse voltage signal in the dq coordinate system is expressed as:
[0068]
[0069] where, u dh and u qh are the injected voltage signals on the d-axis and q-axis respectively, u inj is the voltage signal with different pulse sequences after random selection, T p1 is the injection period of the pulse voltage signal, u s_x (x = a, b, c, d) is the sequence sample of the injected pulse signal, t is the time, and n is the number of periods of the injected pulse voltage.
[0070] The sequence sample of the injected pulse signal can be expressed as:
[0071] u s_x (t, T p1 ) = R[u s_a (t, T p1 ), u s_b (t, T p1 ), u s_c (t, T p1 ), us_d (t, T p1 )]
[0072] Among them, R[] is a random operator, and the pulse voltage signals with sequences a, b, c, and d are respectively expressed as:
[0073]
[0074] u s_b (t, T p1 ) = -u s_a (t, T p1 )
[0075]
[0076] u s_d (t, T p1 ) = -u s_c (t, T p1 )
[0077] Among them, t r (t, T p1) is the remainder of t divided by T p1 , and V inj is the amplitude of the injected voltage. Even if the pulse signals with sequences c or d are injected arbitrarily, the estimated d-axis current will not have a DC offset. However, when the injection probabilities of the pulse signals with sequences a and b are not equal or have a non-ideal mathematical expectation due to non-ideal factors, the estimated d-axis current may have a DC offset. To avoid this problem, the pulse voltage signals of sequences a and b are injected in pairs. Define T p2 = 2T p1 , then the sequence sample of the injected pulse voltage signal can be rewritten as:
[0078] u s_x (t, T pi ) = R[u s_c (t, T p1 ), u s_d (t, T p1 ), u s_c (t, T p2 ) / 2, u s_d (t, T p2 ) / 2]
[0079] Here, i takes 1 or 2. That is, the frequency range of the injected pulse voltage signal is f p1 (= 1 / T p1) to f p2 (= 1 / T p2 ).
[0080] Step 2: After injecting a pulse voltage signal with a random sequence into the motor stator, the permanent magnet synchronous motor generates a high-frequency current response with the same random sequence. The three-phase current of the permanent magnet synchronous motor is obtained through A / D sampling, and the three-phase current is transformed from the natural coordinate system to the stationary coordinate system through Clark transformation:
[0081]
[0082] Step 3: Separate the fundamental wave signal and the high-frequency response signal through a signal demodulation strategy, and extract the envelope of the high-frequency response signal to obtain the orthogonal signal related to the rotor position.
[0083] In this embodiment, since the injection signal frequency is much higher than the motor operating frequency, it can be considered that the fundamental wave current remains unchanged within one injection period. By using the subtraction operation of the two current differences at adjacent signal demodulation instants to achieve signal demodulation, first design the corresponding demodulation signal sym according to different sequences of the pulse signal: αβ sym
[0084]
[0085] sym s_x (t,T pi )=R[sym s_c (t,T p1 ),sym s_d (t,T p1 ),sym s_c (t,T p2 ),sym s_d (t,T p2 )]
[0086] Here, the demodulation signals sym s_c (t,T pi) and sym s_d (t,T pi) can be expressed as:
[0087]
[0088] sym s_d (t,T pi )=-sym s_c (t,T pi )
[0089] The estimated high-frequency response current on the d-axis excited by the injected pulse signal can be expressed as:
[0090]
[0091] i s_x (t,T pi )=R[is_a (t, T p1 ), i s_b (t, T p1 ), i s_a (t, T p2 ), i s_b (t, T p2 )]
[0092] Here, i dh is the d-axis current response signal, and i inj is the current response signal excited by voltage signals with different pulse sequences after random selection. The current response sample can be expressed as: s_x (x = c, d) is the sequence sample of the injected current response signal. The current response samples can be expressed as:
[0093]
[0094] i s_d (t, T pi ) = -i s_c (t, T pi )
[0095] Here, k I is the amplitude of the current response signal. The envelope of the high-frequency current response in the stationary coordinate system can be extracted:
[0096]
[0097] Here, Δi cos and Δi sin are orthogonal signals containing position information. i α,β0 , i α,β1 and i α,β2 are the current sampling values at the signal demodulation times t r (t, T pi) is 0, T pi / 6, and T pi / 3, or the signal demodulation time t r (t, T pi) is T pi / 2, 2T pi / 3, and 5T pi / 6. ΔT is the signal demodulation period, and η is a variable that can be expressed as
[0098] Since the signal demodulation is performed 1 or 2 times in one injection period, a low-pass filter with a high cut-off frequency is used to implement the difference of the orthogonal signals within the control period. Here, the low-pass filter is high enough not to cause phase changes in the orthogonal signals. The filtered orthogonal signals Δi cos1 and Δi sin1 can be expressed as:
[0099]
[0100] After normalization, the unit orthogonal signals containing rotor position information can be expressed as:
[0101]
[0102] Step 4: Use an orthogonal phase-locked loop to process the orthogonal signals related to the rotor position and then obtain the estimated rotor speed of the motor and position
[0103] In this embodiment, after obtaining the unit orthogonal signals containing rotor position information, the equivalent position error ε is obtained by the contrast method:
[0104]
[0105] Furthermore, the input position error is adjusted to 0 through a PI regulator, so that the estimated motor speed and position converge to the actual values, and the estimated motor speed and position can be expressed as
[0106]
[0107] Here, k p and k i are the proportional and integral gains of the phase-locked loop, respectively.
[0108] Step 5: Use the estimated motor and position for speed closed-loop control and Park and inverse Park coordinate transformations respectively to achieve sensorless closed-loop control of the permanent magnet synchronous motor.
[0109] Specific embodiment: To further verify the beneficial effects of the present invention, a specific embodiment is described below.
[0110] The beneficial effects of the present invention are verified through a 750W permanent magnet synchronous motor experimental platform. The parameters of the 750W permanent magnet synchronous motor are: rated frequency 250Hz, rated speed 3000rpm, number of pole pairs 5, d-axis inductance 4.2mH, q-axis inductance 6.2mH, stator resistance 0.9Ω, rated current 4.2A. A hysteresis brake is used as the motor load, a two-level three-phase inverter is used to drive the permanent magnet synchronous motor, and a DSP TMS320F28335 is used to implement the control algorithm. The sampling frequency and the switching frequency are both 10kHz. The injection voltage amplitude is 48V, and T p1 and T p2They are 600 μs and 1200 μs. To ensure that the high-frequency current responses of the excitations have the same amplitude, the voltage-to-frequency ratio of the injected non-zero voltage signal is set to a constant. The amplitudes and voltages of the traditional high-frequency square-wave voltage injection method compared with the method provided by the present invention are set to: 2.5 kHz and 48 V, 1.25 kHz and 24 V, and the amplitudes and voltages of the traditional high-frequency pulse voltage injection method are set to: 1.67 kHz and 48 V.
[0111] Figure 4 It is the comparative analysis diagram of the FFT and PSD of the phase-a current of the present invention method and the traditional method when the permanent magnet synchronous motor operates at 300 rpm and rated load conditions. As Figure 4 shown in (a), (b), and (c) therein, there are high-peak discrete components in the spectral analysis results of the traditional high-frequency square-wave voltage signal injection method and the high-frequency pulse voltage signal injection method, and the frequencies where the discrete components are located are related to the injection frequencies. As Figure 4 shown in (d), the method provided by the present invention can eliminate all discrete harmonics in the spectral analysis results, the spectrum is smooth and continuous, approximately white noise, and the peak value of the current response spectrum can be reduced by about 25 dB compared with the traditional method. Therefore, the method provided by the present invention can suppress the audible noise problem of the sensorless control method based on the signal injection method.
[0112] Figure 5 It is the signal waveform diagram of the signal demodulation method of the present invention method under the conditions of 150 rpm and rated load of the permanent magnet synchronous motor. As shown in the figure, through the signal demodulation method proposed by the present invention, the envelope of the excited high-frequency current response can be extracted, and the extracted unit orthogonal signal containing position information has extremely high sinusoidality, and the estimated rotor position error is within ±5°. Therefore, the signal demodulation strategy adopted by the present invention can effectively filter out interference signals so that the estimated rotor position has high accuracy.
[0113] Figure 6 It is the experimental comparison result of the traditional square-wave injection method and the sensorless method provided by the present invention under the conditions of 300 rpm and half-rated load of the permanent magnet synchronous motor. As shown in the figure, compared with the traditional square-wave injection method, the unit orthogonal signal containing position information extracted by the sensorless method provided by the present invention has a higher signal-to-noise ratio, so that the position estimation accuracy can be improved by about 3.5°, indicating that the sensorless method provided by the present invention has higher observation accuracy than the traditional method.
[0114] Figure 7It is the experimental result of the acceleration and deceleration conditions under the rated load of the method of the present invention. As shown in the figure, during the whole acceleration and deceleration process, the estimated speed and position of the method of the present invention can accurately track the actual motor speed and position information, and the speed and position errors can be ensured within ±10.2 rpm and 12.5°, indicating that the sensorless method provided by the present invention can operate well under the acceleration and deceleration conditions.
[0115] Figure 8 It is the experimental result of the forward and reverse running conditions under the half-rated load of the method of the present invention. As shown in the figure, during the whole forward and reverse running process, the peak value of the estimated position error is within 13.4°, indicating that the sensorless method provided by the present invention can operate well under both forward and reverse running conditions.
[0116] Figure 9 It is the experimental result of the rated load step condition when the speed of the method of the present invention is 300 rpm. The experimental result shows that when the motor is disturbed by the rated load step of the noise, the peak value of the estimated position error is within 12.3°, indicating that the sensorless method provided by the present invention has good dynamic performance and can resist the rated load step disturbance.
[0117] In summary, the sensorless method provided by the present invention has good ability to suppress audible noise and electromagnetic interference, and at the same time can provide good dynamic and static performance. It can broaden the application field of the sensorless control method based on the high-frequency signal injection method, and has important theoretical significance and practical value.
[0118] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited by the above embodiments. The above embodiments and the descriptions in the specification only illustrate the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention will have various changes and improvements, and these changes and improvements all fall within the scope of the present invention claimed.
Claims
1. A sensorless control method for a permanent magnet synchronous motor with random pulse sequence injection, the steps are as follows: Step 1: Use a random pulse sequence voltage signal generator to superimpose a voltage signal with a random pulse sequence on the output of the d-axis current loop estimated by the vector control of the permanent magnet synchronous motor; Step 2: Obtain the three-phase current of the permanent magnet synchronous motor through A / D sampling, and convert the three-phase current from the natural coordinate system to the stationary coordinate system through Clark transformation; Step 3: Separate the fundamental wave signal and the high-frequency response signal through a signal demodulation strategy, and extract the envelope of the high-frequency response signal to obtain an orthogonal signal related to the rotor position; Step 4: Process the orthogonal signals related to the rotor position using an orthogonal phase-locked loop to obtain the estimated rotor speed of the motor with position Step 5: The estimated motor and the position are respectively used for speed closed-loop control and Park and inverse Park coordinate transformations to realize sensorless closed-loop control of the permanent magnet synchronous motor; The pulse sequence of injecting a pulse voltage in Step 1 includes: Sequence a: "+ + - - 0 0", sequence b: "- - + + 0 0", sequence c: "+ - 0 - + 0" and sequence d: "- + 0 + - 0", here, the four different pulse sequences are all composed of six states, each state lasts for one PWM carrier period, "+" represents positive voltage injection, "-" represents negative voltage injection, and "0" represents zero voltage injection; The injected pulse voltage signal in the dq coordinate system is expressed as: Where, udh and uqh are the injected voltage signals on the d-axis and q-axis respectively, uinj is the voltage signal with different pulse sequences after random selection, Tp1 is the injection period of the pulse voltage signal, us_x (x = a, b, c, d) is the sequence sample of the injected pulse signal, t is time, and n is the number of periods of the injected pulse voltage; The sequence sample of the injected pulse signal can be expressed as: u s_x (t,T p1 ) = R[u s_a (t,T p1 ), u s_b (t,T p1 ), u s_c (t,T p1 ), u s_d (t,T p1 )] Where, R[] is a random operator, and the pulse voltage signals with sequences a, b, c, and d are respectively expressed as: u s_b (t, T p1 ) = -u s_a (t, T p1 ) u s_d (t, T p1 ) = -u s_c (t, T p1 ) Where, tr(t, Tp1) is the remainder of t divided by Tp1, Vinj is the amplitude of the injected voltage. Even if the pulse signals with sequences c or d are injected arbitrarily, the estimated d-axis current will not have a DC offset. However, when the injection probabilities of the pulse signals with sequences a and b are not equal or have a non-ideal mathematical expectation due to non-ideal factors, the estimated d-axis current may have a DC offset. To avoid this problem, the pulse voltage signals of sequence a and sequence b are injected in pairs. Define Tp2 = 2Tp1, then the sequence sample of the injected pulse voltage signal can be rewritten as: u s_x (t, T pi ) = R[u s_c (t, T p1 ), u s_d (t, T p1 ), u s_c (t, T p2 ) / 2, u s_d (t, T p2 ) / 2] Here, i takes 1 or 2, that is, the frequency range of the injected pulse voltage signal is from fp1 (= 1 / Tp1) to fp2 (= 1 / Tp2).
2. The sensorless control method for a permanent magnet synchronous motor with random pulse sequence injection according to claim 1, characterized in that, The sequence of the injected pulse voltage signal is randomly selected and generated by a random pulse sequence voltage signal generator.
3. The sensorless control method for a permanent magnet synchronous motor with random pulse sequence injection according to claim 1, wherein The specific signal demodulation method in step 3 is as follows: First, obtain the three-phase current of the permanent magnet synchronous motor through A / D sampling, and convert the three-phase current from the natural coordinate system to the stationary coordinate system through Clark transformation. Since the injection signal frequency is much higher than the motor operating frequency, it can be considered that the fundamental wave current remains unchanged within one injection period. The signal demodulation is realized by subtracting the difference between two adjacent currents at the moment of iαβ signal demodulation. First, design the corresponding demodulation signal sym according to different sequences of the pulse signal: sym s_x (t, T pi ) = R[sym s_c (t, T p1 ), sym s_d (t, T p1 ), sym s_c (t, T p2 ), sym s_d (t, T p2 )] Here, the demodulation signals syms_c(t, Tpi) and syms_d(t, Tpi) can be expressed as: sym s_d (t, T pi ) = -sym s_c (t, T pi ) The estimated high-frequency response current on the d-axis excited by the injection pulse signal can be expressed as: i s_x (t, T pi ) = R[i s_a (t, T p1 ), i s_b (t, T p1 ), i s_a (t, T p2 ), i s_b (t, T p2 )] Here, idh is the d-axis current response signal, iinj is the current response signal excited by the voltage signal with different pulse sequences after random selection, is_x (x = c, d) is the sequence sample of the injection current response signal, and the current response sample can be expressed as: i s_d (t, T pi ) = -i s_c (t, T pi ) Here, kI is the amplitude of the current response signal and satisfies (kI = Vinj·Tp1 / 6Ld), and the envelope of the high-frequency current response in the stationary coordinate system can be extracted: Here, Δicos and Δisin are orthogonal signals containing position information, iα,β0, iα,β1, and iα,β2 are the current sampling values at the signal demodulation times tr(t, Tpi) being 0, Tpi / 6, and Tpi / 3, or at the signal demodulation times tr(t, Tpi) being Tpi / 2, 2Tpi / 3, and 5Tpi / 6, ΔT is the signal demodulation period, and η is a variable that can be expressed as Since the signal demodulation is performed 1 or 2 times in one injection period, a low-pass filter with a high cut-off frequency is used to realize the difference of the orthogonal signals within the control period. Here, the low-pass filter is high enough not to cause phase changes in the orthogonal signals. The filtered orthogonal signals Δicos1 and Δisin1 can be expressed as: After normalization, the unit orthogonal signal containing the rotor position information can be expressed as:
4. The sensorless control method for a permanent magnet synchronous motor with random pulse sequence injection according to claim 1, wherein The equivalent position error ε is obtained by the contrast method as described above: Furthermore, the input position error is adjusted to 0 through a PI regulator, so that the estimated motor speed and position converge to the actual values. The estimated motor speed and position can be expressed as Here, kp and ki are the proportional and integral gains of the phase-locked loop respectively.
Citation Information
Patent Citations
Permanent magnet synchronous motor vibration and noise reduction method based on harmonic selective elimination random SVPWM
CN112104281A
Method and apparatus for estimating rotor position of electric motor, and electric motor control system
WO2020155864A1